AWS Weekly Roundup: DuckLabs Acquisition and Latest Service Updates
- Amazon Web Services has signed a definitive agreement to acquire Amsterdam-based DuckLabs, the company behind the open-source analytical database DuckDB, according to recent tech industry disclosures.
- DuckDB will remain open source under an independent foundation and the MIT license.
- Co-founded by Hannes Mühleisen and Mark Raasveldt, DuckDB executes SQL directly against file formats including Parquet, CSV, and JSON either locally or on Amazon S3.
AWS Acquires DuckDB Creator DuckLabs
Amazon Web Services has signed a definitive agreement to acquire Amsterdam-based DuckLabs, the company behind the open-source analytical database DuckDB, according to recent tech industry disclosures.
DuckDB will remain open source under an independent foundation and the MIT license. Meanwhile, AWS plans to integrate its in-process query speed with enterprise services like Amazon S3, Amazon Redshift, and Amazon Athena.
Co-founded by Hannes Mühleisen and Mark Raasveldt, DuckDB executes SQL directly against file formats including Parquet, CSV, and JSON either locally or on Amazon S3. According to verified industry reports, the database is optimized for everyday analytical queries involving a terabyte or less of data. The architecture also pairs naturally with AI agents that experiment and query information dynamically.
Following the acquisition, the co-founders will continue directing the technical roadmap. At the same time, AWS combines DuckDB capabilities with existing platforms like Amazon EMR, AWS Glue, and Amazon SageMaker. Andy Warfield, Vice President and Distinguished Engineer at AWS, outlined the broader architectural implications in a post titled DuckDB and the changing physics of analytics on All Things Distributed.
Container Recovery and Lambda Previews
Amazon ECS now automatically detects and recovers container instances that lose agent connectivity. The control plane continuously monitors agent status and surfaces a new AGENT_CONNECTIVITY health event across AWS Fargate, Amazon ECS Managed Instances, and Amazon ECS on EC2.
For Fargate and Managed Instances, the service manages recovery automatically by draining tasks, launching replacements, and deregistering impaired instances at no additional cost across all AWS Commercial and AWS GovCloud Regions.
Additionally, AWS Lambda has introduced public preview runtimes, starting with Node.js 26 and Python 3.15. Developers can test upcoming runtimes ahead of general availability using identifiers that match the eventual production versions, allowing functions to graduate automatically. These preview runtimes are available across all commercial, AWS GovCloud (US), and China Regions, though breaking changes remain possible prior to general availability.
Native Time-Series Routing via IoT Core
AWS IoT Core has added a native InfluxDB rule action. This allows organizations to route time-series device data directly into managed or self-hosted InfluxDB instances without custom code or intermediate services.

The integration formats data into InfluxDB line protocol and supports both device-side and server-side batching wherever Amazon Timestream for InfluxDB is available.
Shielded Game Servers and SageMaker Ray Clusters
Amazon GameLift Servers has rolled out enhanced DDoS protection for network and transport layers, including UDP reflection and SYN floods. Built on AWS Shield Standard with gaming-optimized traffic shaping using Server SDK 5, the protection activates automatically at no extra cost in all supported regions except China (Beijing) and China (Ningxia).
Meanwhile, Amazon SageMaker HyperPod has expanded support for Ray workloads. This enables multi-node cluster management directly from Amazon SageMaker Studio with auto-provisioned Grafana dashboards, node auto-recovery, and tiered checkpointing.
